Adult tissue multipotent stem cells have a double regulatory program: one that preserves their ability to differentiate into various cell types, and one that allows them to self-renew while preventing or postponing differentiation. For the cells to differentiate, the link between these two program components must be broken. Differentiation vs. self-renewal decisions must involve activation and cross-inhibition of different circuits in a gene regulatory network, but these network architectures are not well understood. It is not known at a molecular level how individual multipotent stem cells control the timing of their decisions to differentiate, or the speed with which they will differentiate relative to clonal expansion to a given fate. This is the problem that we propose to address using a combination of single-cell live imaging, gene network analysis, and computational modeling of population dynamics and gene network dynamics in a particularly tractable experimental system. In hematopoiesis, this wider problem can be addressed because there is excellent characterization of short- term and long-term multipotent stem cells, plus a variety of partially restricted progenitor cells which have different repertoires of developmental potential but still defer lineag commitment, keeping multiple options open. The T-lymphocyte developmental pathway is one branch of hematopoiesis in which this general question may be unusually accessible. It is based on a well-defined series of intermediate states with reproducible developmental and gene expression characteristics, and the T cell program can be triggered and guided in hematopoietic precursors experimentally in vitro. At the same time, the path to T-cell lineage commitment involves extensive proliferation, raising the question of how proliferation that advances differentiation may be distinguished from self-renewal, and providing a system in which to test factors that control this distinction. We have assembled a multidisciplinary systems biology collaboration to dissect the basis of the choice to enter the T-cell pathway, by an integrated strategy of computational modeling and experimental analysis. Our initial work has shown that T cell precursors initially go through a phase of self-renewal-like proliferation in which differentiation competence is delayed, and that readiness to differentiate in single cells is finaly provided by intrinsic regulatory changes. The regulatory circuitry underlying this switch is now accessible by exploiting new access to single-cell live tracking and gene expression analysis methods while leveraging the results of our recent genome-wide transcriptome analyses of early T-cell precursors. Here we propose: to establish a framework for the problem by modeling the kinetics of T-lineage differentiation choices relative to proliferation in individual cells; to trak individual cell fates forward and backward through the T-cell developmental process by live imaging; to determine gene expression features of individual cells that best predict their developmental behavior; and to use computational and experimental approaches in an interlaced, iterative way to deduce the gene network architecture that controls entry into the T-cell pathway, and to validate its key linkages.